Sunday, September 27, 2009

Generations and Jeopardy! wagering tendencies

After several name and gender related blog posts, I'm finally going to move onto another subject (the S&H theory will still be involved though). One of my favorite game shows is Jeopardy!, and I've done some studying on wagering theory in that game. I'm posting this blog discussing how general wagering habits of the contestants have changed over time.

Jeopardy! fans who remember watching the show during the Art Fleming era may remember that back then all three contestants won whatever money they had (but like now only the winner got to return to play again). When the Trebek version started in 1984 the rules changed so that only the winner got to keep his/her winnings, mainly because back in the Fleming days contestants would often give up any chance at winning so that he/she could keep whatever had been won at that point. The rule changed helped make the game more competitive since the score would now effectively be only points until the game was won. I have a good explanation why Fleming-era contestants tended to be more cautious with their winnings, and why some people think the game should return to the original format of everyone keeping what they have. The secret is in the generational archetype of the contestants. Of course what I'm about to say is a very broad generalization and you will very likely find contestants who don't fit the general description, but in terms of general tendencies it's a good description.

According to William Strauss and Neil Howe there are four different "archetypes" of generations with a different one succeeding each other in a fixed order and repeating every four generations, with each generation (in modern times) lasting about 20 years or so (an exception was around the time of the Civil War in which one of the archetypes was skipped, for reasons I won't get into here). Each archetype varies in several attributes, such as whether they're individually or collectively focused, or whether they're risk-adverse or risk-takers. The particular attributes I gave examples to are the ones that give some clues to how one would wager on Jeopardy!.

In Jeopardy!-fan speak, "Venusian" refers to one who likes to wager small, i.e. risk-adverse and "Martian" refers to the opposite, i.e. one who likes to wager big. In terms of generational archetypes, there is one that tends to be risk-adverse (the Adaptives or Artists) and one on the opposite side of the saeculum (the 80-year or so cycle of generations and types of eras) that tends to be risk-takers (the Reactives or Nomads). Examples of the former include the Silent Generation (which S&H define as those born from 1925-1942) and the current crop of children being born (since the early to mid years of this decade). Examples of the latter are Generation X (born 1961-1981) and before them the Lost Generation (born 1883-1900). The generations in between (which I'll get into more detail on later in this blog) will tend to be somewhere between these two extremes. During the orignal Fleming era (1964-1975) Silent contestants probably made up a plurality of the total players of the time. As I mentioned Silents come from a risk-adverse generation, and as you might expect that would apply to Jeopardy! as well (hence the large number of contestants from that era who would rather "keep what they had" then to try and go for the win if they could). The opposite types of generations were largely out of the picture at the time (there were probably at least a few Lost contestants but probably not a significant number, and unless they had a Kids or [at the very end of the era] Teen version Xers would've all been too young to try out).

At the start of the Trebek era Silents probably no longer made up the plurality of contestants (that honor went to the next generation, the Boomers) but still a singnificant number, hence the implementation of the "winner take all" rule. Also at the time Xers were just starting to age into the young contestant range, and later on likely took the "plurality of contestants" honor. This means that now, unlike during the Fleming era, there are more contestants from a risk-taking generation (Xers) than a risk-adverse one (although still a few Silents not as many as before, and any members of the Homeland generation [the tentative name for the generation which is being born into right now] are too young even for Kid's Week). If the rules were to revert to the keep-what-you-have setup of the Fleming era, I think that the number of Venusian contestants would be lower than back then.

Now on to the other two types of generations. An attribute of Idealist or Prophet generations (e.g. Boomers, born between 1943-1960) is that they tend to be individualistic (hence phrases like the "Me Generation"). In terms of Jeopardy! wagering that means that they are less likely to wager for a tie for "generosity" reasons, with the opposite applying to the collectivist Civic or Hero generations (e.g. the Millennials born between 1982 and sometime in the early to middle years of the current decade, and before them the G.I.s born 1901-1924). Alternate terms that are more widely known to Jeopardy! fans that may be used to describe these various behaviors include "debunkitive" and "cooperative". This logic may also be used (although with no certainty) to predict whether or not players in a Prisoner's Dilemma (where the leaders are tied and the third contestant if there is one is too distant to catch up to them if they wager nothing) will "cooperate" by betting nothing and ensuring a co-win as long as they both follow through with the wagering or "debunk" by betting everything to ensure a win (or co-win) as long as the contestant in question gets Final Jeopardy! right (note that $0 is never considered a winning score on Jeopardy!, so it is impossible to win if you bet everything and respond to Final Jeopardy! wrong). With the Millennials being the fastst growing group of contestants, I predict that there may be more contestants betting for the tie in the coming years (incidentally restoring the old Fleming-era setup mentioned earlier may reduce this, since there won't be an incentive to tie so that The Powers That Be end up paying out more to the contestants).